444 Mr. E. L. Garbett's Description of some Parhelia 



If all the icy prisms contained in the air have their axes 

 truly parallel, it is plain that these phaenomena will be unac- 

 companied by any ordinary halo ; as in the first of the displays 

 above described : but if a portion of the crystals be dispersed 

 by the wind, or a cessation of the cause (electrical induction, 

 or whatever it might be) that kept them in their polar state, 

 the portion so scattered would, by refracting in all directions, 

 produce a circular halo in addition to the former appearances; 

 as actually happened in the afternoon display : but as long as 

 this scattering was not complete, so as to turn the prisms in 

 all directions equally, the halo would not be equally luminous 

 throughout, but as shown in fig. 3. On the following day, 

 however, all vestiges of a polar arrangement had disappeared, 

 and the refraction being performed eqiialli) in all directions, 

 produced a regular halo of greatly increased splendour, but 

 without any of the other appearances. 



We have as yet spoken only of the light reflected and re- 

 fracted by the sides of the icy prisms ; but it is plain that their 

 ends, whether flat, or terminated by pyramids or any other 

 forms, must produce effects of a similar kind, but more com- 

 plicated ; and to their action we must refer the various other 

 arcs and parhelia described and figured above, and differently 

 in every different account of such phaenomena. If we con- 

 sider in how many different ways a prismatic crystal may be 

 terminated, from a flat face to the most complex pyramids, or 

 the tables and still more complex forms observed in snow and 

 rime, we shall indeed find no difficulty in believing, that hardly 

 any two tlisplays of parhelia agree in the form and arrange- 

 ment of their circles and arcs, except so far as regards the 

 white horizontal circle and the two spectra with tails, on the 

 right and left of the sun, which seem to be common to all 

 accounts of these phaenomena: as they must be if due, as we 

 have suggested, to the action of the invariable sides of the 

 prisms, independently of their ends. 



It would doubtless be possible, from the description of one 

 of these exhibitions, to determine the precise form of both the 

 upper and lower ends of the icy crystals which produced it; 

 or, conversely, to determine what appearances would be pro- 

 duced in an atmosphere charged with crystals of a given form. 

 Both these problems, however, would be very laborious ; but 

 there is a way of solving the second of them by experiment, 

 which I think will serve to show very clearly the rationale of 

 these ap}iearances. 



Suppose that in a dark room we have a large white hemi- 

 spherical surface, placed like a basin, with its edge horizontal, 

 and in the centre of its concavity we suspend from above a 

 small prismatic crystal of any kind, with its axis vertical. This 



